Loading...

Semi-automated Distribution System Planning with Considering Decentralized Automation at Feeder Level

Fallah, Ali | 2022

163 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 55436 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Fotuhi Firuzabad, Mahmoud
  7. Abstract:
  8. Implementation of telecommunication-based distribution automation is possible in both centralized and decentralized manners. Although centralized automation is more pervasive than decentralized automation, benefits of decentralized automation systems, like improving reliability and speeding up the fault management process, encourage distribution companies to implement these systems. However, high cost of decentralized automation can be a deterrent to most distribution companies. Therefore, economic planning and creating economic justifications for decentralized distribution automation are necessary for the development of these systems. In this dissertation, economic analysis and planning of decentralized distribution automation at MV feeder level have been done. In this regard, first, implementation of decentralized automation in both distributed and hierarchical methods have been investigated based on the concept of multi-agent systems. Then, cost-benefit analysis of equipping distribution system with decentralized automation is performed considering reliability improvement. According to these studies, decentralized automation planning has been done in two phases from the perspective of social welfare and distribution companies. In the first phase, automation planning is implemented to maximize the social welfare, according to the value of lost load. The goal of planning was to place the agents of the multi-agent system optimally. Genetic algorithm has been used to solve the optimization problem. Results of the simulations have shown the significant economic justification of decentralized automation in terms of social welfare. Furthermore, according to the different scenarios defined for the initial state of the network, the dependence of the optimal location of the agents and the switching equipment in the network has been determined. Therefore, the need for integrated switch and agent placement at MV feeder level to achieve optimal network development plan has been concluded.In the second phase, automation planning is done from the perspective of distribution companies. According to the results of the first phase, an integrated switch and agent placement planning with more economic justification for distribution companies has been presented. The problem is modeled as five-year multi-stage planning. Due to the complexity and high number of possible planning scenarios, an innovative method has been developed to solve the optimization problem by combining genetic algorithm and simulated annealing. The results have shown that the revenue increasement of distribution companies due to selling more energy resulted from reliability improvement is insignificant compared to network development costs. Therefore, investments have been made to the extent that the constraints of the problem are met, including minimum acceptable reliability indices.In this dissertation, the economic benefits of decentralized distribution automation planning in the form of multi-agent systems are studied. According to the simulations, the effectiveness of switching devices on agent placement in the distribution network has been shown. Thus, an integrated switch and agent placement have been planned from the perspective of distribution companies. To solve this problem, a hybrid GA/SA algorithm has been presented. The results of this dissertation can be used by distribution companies and would encourage them to implement multi-agent systems at feeder automation. To make this study more comprehensive, in addition to the network fault management improvement, other benefits of decentralized automation can be considered in future studies
  9. Keywords:
  10. Multiagent System ; Feeder Automation ; Cost-Benefit Analysis ; Meta Heuristic Algorithm ; Decentralized Automation ; Automation Planning ; Semi Automatic System ; Distribution Networks Planning

 Digital Object List

 Bookmark

No TOC